Expensya Founders Return with Thunder Code, an AI Software Testing Startup, Raising $9M Seed

From Major Exit to AI Frontier: Expensya Founders Launch Thunder Code with $9M Seed

In the dynamic world of technology and entrepreneurship, stories of successful exits often mark the culmination of years of hard work, innovation, and resilience. For Karim Jouini and Jihed Othmani, the co-founders of Expensya, a leading expense management startup, their journey reached a significant milestone in 2023. They successfully sold Expensya to Swedish procurement software firm Medius in a deal widely recognized as one of the largest acquisitions of an African startup to date. While the exact terms of the deal were not publicly disclosed, some reports placed the acquisition sum north of $120 million, a figure that underscores the scale of Expensya's success and its impact on the African tech ecosystem. The acquisition also saw former and current Expensya employees cash out a reported $10 million, a testament to the value created and shared within the company.

Following this momentous exit, both Jouini and Othmani expressed a clear intention to step back from the demanding world of founding startups. Having navigated the complexities of building and scaling a company across multiple continents, the desire for a less intense path was understandable. Jouini transitioned into a Chief Technology Officer role within the newly merged entity at Medius, overseeing the technical integration of several acquired companies across Europe, North America, and Africa. This period offered a different perspective, moving from the intense, all-consuming focus of a founder to a strategic leadership position within a larger corporate structure.

The Irresistible Pull of Generative AI

Despite their initial resolution to avoid the startup grind, a powerful new technological wave began to crest: generative AI. The rapid advancements and transformative potential of this technology proved too compelling for the seasoned entrepreneurs to ignore. As Jouini immersed himself in the technological landscape at Medius, integrating diverse software systems, a universal challenge became glaringly apparent: software testing. Regardless of the product, the team, or the geography, the process of ensuring software quality was consistently slow, resource-intensive, and often a bottleneck in the development lifecycle.

This firsthand experience, combined with Othmani's deep background in artificial intelligence – having built internal AI tools at Expensya years before generative AI became mainstream – sparked a new idea. They realized that generative AI wasn't just a tool for content creation or data analysis; it held immense promise for fundamentally reshaping how software is built and validated. The thought began to take root: could they leverage this new wave to build something even bigger, something that addressed a core pain point across the entire software industry?

This realization led them back to the entrepreneurial arena, driven by the conviction that they could apply their hard-won experience and Othmani's AI expertise to tackle the pervasive problem of software testing. The initial reluctance faded, replaced by the familiar fire of building something disruptive. As Jouini put it, returning to the startup world after Expensya was like having a second child – you forget how hard the first one was, but the excitement and potential make the intensity worthwhile.

Introducing Thunder Code: AI-Powered Software Testing

Their new venture, officially launched less than six months ago, is called Thunder Code. The company is building a generative AI-powered platform designed to revolutionize software testing. At its core, Thunder Code utilizes sophisticated AI agents that are trained to mimic the actions and thought processes of human quality assurance (QA) testers. These agents are capable of autonomously navigating software applications, simulating user interactions, identifying bugs, and even pinpointing subtle user interface (UI) and user experience (UX) issues that traditional automated tests might miss.

The traditional software testing paradigm often relies heavily on manual testing or brittle, code-based automation scripts that require constant maintenance. This leads to slow feedback loops, delayed releases, and significant costs. Thunder Code aims to break this cycle by offering a more intelligent, adaptable, and efficient approach. The AI agents can learn from feedback, adapt to changes in the application, and provide comprehensive test coverage much faster than conventional methods.

The platform's ability to simulate human behavior is a key differentiator. While many automation tools focus on verifying specific functionalities based on predefined scripts, Thunder Code's AI agents can explore the application more dynamically, uncovering issues in unexpected user flows or edge cases. This approach promises to significantly accelerate the QA process, allowing development teams to identify and fix issues earlier in the development cycle and ship higher-quality software faster.

Thunder Code team photo — Thunder Code team.Image Credits: Thunder Code

Applying Lessons Learned: Speed, Talent, and Dilution

Having successfully built and exited a major startup, Jouini and Othmani are approaching Thunder Code with valuable insights gained from their Expensya journey. One of the most critical lessons they are applying is the importance of speed and rapid iteration. Jouini emphasized this, stating that they shipped their first Minimum Viable Product (MVP) within just six weeks of starting the company. He contrasted this with Expensya, noting that Thunder Code's product is already significantly more robust and mature after six months than Expensya was after four years. This accelerated pace reflects a core startup principle: getting a working product into the hands of users quickly to gather feedback and iterate, rather than spending years perfecting a solution in isolation.

Another key lesson is the critical role of talent. The founders are prioritizing hiring top-tier engineers and AI experts from the outset. Jouini is pragmatic about the need for early dilution of equity to attract and retain the best people. This contrasts with a common tendency among some entrepreneurs, particularly in emerging ecosystems, to hold onto a larger percentage of equity even if it means compromising on talent. Jouini and Othmani believe that building a truly valuable, potentially 'unicorn' company requires bringing in the best minds, and they are willing to dilute their ownership stake to achieve that goal. Their philosophy is that a smaller percentage of a massive success is far more valuable than a larger percentage of a modest one.

Furthermore, the founders are focusing intensely on the core features that deliver the most value to their initial customers. This disciplined approach, honed at Expensya, helps avoid feature creep and ensures that the product solves the most pressing problems for their target market effectively. By concentrating their resources and efforts, they can build a strong foundation and achieve product-market fit faster.

Early Traction and Market Opportunity

Despite its young age, Thunder Code is already demonstrating promising early traction. The company has secured paying customers and is running pilot programs across diverse geographies, including the United States, Canada, France, and Tunisia. Their initial target audience includes delivery managers, dedicated QA shops, and internal developer teams within companies who are eager to accelerate their testing processes and improve software quality.

The choice of software testing as a market is strategic. While it might seem like a departure from expense management, Jouini sees it as a significantly larger and more complex opportunity. The global software testing market is projected to be massive, with estimates suggesting it could exceed $100 billion by 2027. This vast market is still heavily reliant on manual processes and legacy code-based automation tools, many of which may struggle to adapt to the rapid changes brought about by AI.

Thunder Code's current focus is on web application testing, a critical area for most businesses today. However, the company has ambitious plans for expansion. By late 2025, they intend to extend their AI-powered testing capabilities to cover mobile applications, desktop software, and APIs. This planned expansion positions them to address a broader spectrum of software development needs and capture a larger share of the market.

The market for AI-powered software testing is becoming increasingly competitive. Established players like UiPath are exploring agentic testing, and a wave of new startups, such as Jetify and Nova AI, are also entering the space with their own AI-driven solutions. However, Jouini believes Thunder Code's combination of fast execution, deep AI expertise from Othmani, and the application of hard-earned lessons from their previous venture gives them a strong edge in this emerging landscape.

Strong Investor Confidence and Future Outlook

The confidence in Jouini and Othmani's ability to build another successful company is clearly reflected in their initial funding round. Thunder Code has successfully secured $9 million in seed funding within just six months of its inception. This is a substantial amount for a seed round and speaks volumes about the investors' belief in the team and the market opportunity.

The funding round saw participation from several familiar faces from the Expensya cap table, including prominent investors like Silicon Badia and Janngo Capital. Their continued support demonstrates a strong relationship and trust built during the Expensya journey. The round also included Titan Seed Fund and a roster of strategic angel investors, such as Roxanne Varza, the Director of Station F (a major startup campus in Paris), and Karim Beguir, CEO of InstaDeep, which is recognized as Africa's biggest AI startup. This mix of institutional and strategic investors provides Thunder Code with both capital and valuable industry connections.

Perhaps most uniquely, the funding round also includes investments from former and current Expensya employees who benefited from the acquisition. This is a powerful endorsement, indicating that those who worked closest with the founders at their previous company have strong faith in their ability to succeed again. Jouini expressed particular satisfaction in this aspect, highlighting the positive outcome for his former team members.

Thunder Code is headquartered in Paris, leveraging the city's growing tech ecosystem, while maintaining a significant office in Tunis, tapping into the strong technical talent pool in Tunisia, where Expensya was founded. This dual presence allows them to operate in key global markets while retaining a connection to their roots.

Looking ahead, Jouini and Othmani are clearly aiming high. They are convinced that Thunder Code has the potential to become a 'unicorn' – a startup valued at over $1 billion. They believe that the combination of their experience, the transformative power of generative AI, a massive market opportunity, and a focus on speed and talent positions them for significant growth. While the journey of building a startup is never easy, their track record and the early momentum of Thunder Code suggest they are well-equipped to tackle the challenges and potentially achieve even greater success in their second act.